I am using a gravity model to study the impact of cultural differences on the total merger volume from country a to country b (in the 2000 till 2016). My dependent variable is the natural log of the total volume from country a to country b in year t. I want to add time fixed and country fixed effect. Also I want to double cluster the target and acquirer nations standard errors. My regression looks as follows;
ln(Total Volume)= a+ B(Cultural distance)+ ln(distance)+ ln(bilateral trade flow)+ other control variables+ time fixed effect+ target country fixed effects+ acquirer country fixed effects
I have made a panelid with the
egen CrossCountry=group(AcquirerNation TargetNation)
I am using stata and I have set my paneldata as follows;
xtset panelid year
I want to double cluster my standard error on the TargetNation and AcquirerNation to account for within country time-serie correlation. I was wondering if someone could help me out with this.
Kind Regards,
Rik
Dear Rik,
In stata, you can simple use the xtivreg2 command using the cluster option. It should be written as follows : cluster(AcquirerNation TargetNation) and not cluster(CrossCountry).
I also recommand to include a country-pair fixed effect using the fe option after xtset panelid year.
The xtivreg2 command can be used without instrumenting for endogenous variables.
The revelant sources for your question are:
Cameron, A., Gelbach, J., & Miller, D. (2011). Robust Inference With Multiway Clustering. Journal of Business & Economic Statistics, 29(2), 238-249. Retrieved from http://www.jstor.org/stable/25800796
Egger, P. H., & Tarlea, F. (2015). Multi-way clustering estimation of standard errors in gravity models. Economics Letters, 134, 144-147.
Anderson, J. E., & Van Wincoop, E. (2003). Gravity with gravitas: a solution to the border puzzle. American economic review, 93(1), 170-192.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression